agent serialization (#4642)

langchain/agents/agent.py

@@ -12,6 +12,7 @@ from typing import Any, Dict, List, Optional, Sequence, Tuple, Union
 import yaml
 from pydantic import BaseModel, root_validator
 
+from langchain.agents.agent_types import AgentType
 from langchain.agents.tools import InvalidTool
 from langchain.base_language import BaseLanguageModel
 from langchain.callbacks.base import BaseCallbackManager
@@ -132,7 +133,11 @@ class BaseSingleActionAgent(BaseModel):
     def dict(self, **kwargs: Any) -> Dict:
         """Return dictionary representation of agent."""
         _dict = super().dict()
-        _dict["_type"] = str(self._agent_type)
+        _type = self._agent_type
+        if isinstance(_type, AgentType):
+            _dict["_type"] = str(_type.value)
+        else:
+            _dict["_type"] = _type
         return _dict
 
     def save(self, file_path: Union[Path, str]) -> None:
@@ -307,6 +312,12 @@ class LLMSingleActionAgent(BaseSingleActionAgent):
     def input_keys(self) -> List[str]:
         return list(set(self.llm_chain.input_keys) - {"intermediate_steps"})
 
+    def dict(self, **kwargs: Any) -> Dict:
+        """Return dictionary representation of agent."""
+        _dict = super().dict()
+        del _dict["output_parser"]
+        return _dict
+
     def plan(
         self,
         intermediate_steps: List[Tuple[AgentAction, str]],
@@ -376,6 +387,12 @@ class Agent(BaseSingleActionAgent):
     output_parser: AgentOutputParser
     allowed_tools: Optional[List[str]] = None
 
+    def dict(self, **kwargs: Any) -> Dict:
+        """Return dictionary representation of agent."""
+        _dict = super().dict()
+        del _dict["output_parser"]
+        return _dict
+
     def get_allowed_tools(self) -> Optional[List[str]]:
         return self.allowed_tools
 

langchain/agents/loading.py

@@ -1,5 +1,6 @@
 """Functionality for loading agents."""
 import json
+import logging
 from pathlib import Path
 from typing import Any, List, Optional, Union
 
@@ -12,6 +13,8 @@ from langchain.base_language import BaseLanguageModel
 from langchain.chains.loading import load_chain, load_chain_from_config
 from langchain.utilities.loading import try_load_from_hub
 
+logger = logging.getLogger(__file__)
+
 URL_BASE = "https://raw.githubusercontent.com/hwchase17/langchain-hub/master/agents/"
 
 
@@ -61,6 +64,13 @@ def load_agent_from_config(
         config["llm_chain"] = load_chain(config.pop("llm_chain_path"))
     else:
         raise ValueError("One of `llm_chain` and `llm_chain_path` should be specified.")
+    if "output_parser" in config:
+        logger.warning(
+            "Currently loading output parsers on agent is not supported, "
+            "will just use the default one."
+        )
+        del config["output_parser"]
+
     combined_config = {**config, **kwargs}
     return agent_cls(**combined_config)  # type: ignore
 

langchain/llms/__init__.py

@@ -10,6 +10,7 @@ from langchain.llms.base import BaseLLM
 from langchain.llms.cerebriumai import CerebriumAI
 from langchain.llms.cohere import Cohere
 from langchain.llms.deepinfra import DeepInfra
+from langchain.llms.fake import FakeListLLM
 from langchain.llms.forefrontai import ForefrontAI
 from langchain.llms.google_palm import GooglePalm
 from langchain.llms.gooseai import GooseAI
@@ -71,6 +72,7 @@ __all__ = [
     "PredictionGuard",
     "HumanInputLLM",
     "HuggingFaceTextGenInference",
+    "FakeListLLM",
 ]
 
 type_to_cls_dict: Dict[str, Type[BaseLLM]] = {
@@ -105,4 +107,5 @@ type_to_cls_dict: Dict[str, Type[BaseLLM]] = {
     "writer": Writer,
     "rwkv": RWKV,
     "huggingface_textgen_inference": HuggingFaceTextGenInference,
+    "fake-list": FakeListLLM,
 }

langchain/llms/fake.py

@@ -29,4 +29,4 @@ class FakeListLLM(LLM):
 
     @property
     def _identifying_params(self) -> Mapping[str, Any]:
-        return {}
+        return {"responses": self.responses}

tests/unit_tests/agents/test_serialization.py

@@ -0,0 +1,19 @@
+from pathlib import Path
+from tempfile import TemporaryDirectory
+
+from langchain.agents.agent_types import AgentType
+from langchain.agents.initialize import initialize_agent, load_agent
+from langchain.llms.fake import FakeListLLM
+
+
+def test_mrkl_serialization() -> None:
+    agent = initialize_agent(
+        [],
+        FakeListLLM(responses=[]),
+        agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION,
+        verbose=True,
+    )
+    with TemporaryDirectory() as tempdir:
+        file = Path(tempdir) / "agent.json"
+        agent.save_agent(file)
+        load_agent(file)